Replication Archive
"Responsive Campaigning: Evidence from European Parties"

Miguel M. Pereira, WUSTL, m.pereira@wustl.edu
(feel free to reach out if you have questions)

Version: March 11, 2019

The replication file includes:
1) Party-day data: party_day_JOP_Pereira.csv
2) CCDP raw data clean: self_clean.csv
3) Alternative measures of campaign rhetoric: alternative_rhetoric_codings.R
4) Replication file: JOP_Pereira_replication.R
5) All polls used in the study: all_polls.csv


To replicate the results:
a) Open JOP_Pereira_replication.R
b) Set the working directory to location of (1)-(3)
c) Run code


VARIABLES' DESCRIPTION IN Party-day dataset:
camp.day -> Campaign day
party -> party (codes from CCDP)
country -> UK - United Kingdom; SV - Sweden; NL - Netherlands; ES - Spain; HU - Hungary
		   DK - Denmark; DE - Germany; CZ - Czech Republic; PL - Poland; PO - Portugal
year -> campaign year
camp -> paste(country,year,sep='')
party_camp -> paste(party,country,year,sep='')
election_day -> day of upcoming election
daystoelec -> days between campaign day and election day
poll -> 1 - at least one new poll published on this day; 0 - otherwise
share -> poll share
prevvote -> vote share from last general election
elecvote -> vote share in upcoming election
linear_int -> Linear interpolation (see main text for description)
const_int -> Constant interpolation (assigning share of previous poll to days without polls)
dom_avissue_code -> CMP code for dominant issue according to party manifesto (among issues available in CCDP)
dom_avissue_share -> Share of manifesto policy statements on dominant issue
dom_avissue_st -> 1 - Statement(s) made on dominant issue; 0 - otherwise
sec_avissue_st -> 1 - Statement(s) made on second most common issue; 0 - otherwise
left_right -> -1 - left wing; 1 - right wing
vdiff -> linear_int - prevvote (difference between linear interpolation and previous vote share)
vdiffc -> const_int - prevvote (difference between constant interpolation and previous vote share)
poldist -> days since last poll
incumbent -> 1 - member of government coalition; 0 - otherwise
pm -> 1 - controls executive; 0 - otherwise
junior -> 1 - junior coalition partner (no control of PM); 0 - otherwise
enp -> effective number of parties
coh_1day -> # ideologically consistent - # ideologically inconsistent policy statements on this day
cohthey_0day -> # ideologically consistent - # ideologically inconsistent policy statements on this day made by other parties
coh_1day_lag -> # ideologically consistent - # ideologically inconsistent policy statements on the day before
niche -> 1 - niche party; 0 - mainstream party
propdayscovered -> share of campaign days with statements from or about the party
cons -> 1 - net ideologically consistent set of policy statements ; 0 - net ideologically inconsistent set of policy statements on this day
cons_lag -> 1 - net ideologically consistent set of policy statements ; 0 - net ideologically inconsistent set of policy statements on the day before
any_ideost -> # of ideological statements on a given day
cons3d -> 1 - net ideologically consistent set of policy statements ; 0 - net ideologically inconsistent set of policy statements on the following 3 days
cons3d_lag -> 1 - net ideologically consistent set of policy statements ; 0 - net ideologically inconsistent set of policy statements on the following 3 days
niche_alt -> 1 - niche party; 0 - mainstream party following Meguid (2005) typology
chal -> 1 - challenger; 0 - otherwise, following De Vries and Hobolt (2012)
vdiff_2yb -> current interpolated share - interpolated share 2 years before
vdiff_1yb -> current interpolated share - interpolated share 1 year before
vdiff_2mb -> current interpolated share - interpolated share 2 months before
vdiff_1mb -> current interpolated share - interpolated share 1 month before
vdiff_1wb -> current interpolated share - interpolated share 1 week before
vdiff_1db -> current interpolated share - interpolated share 1 day before
green -> 1 - green party; 0 - otherwise
rrw -> 1 - radical right wing party; 0 - otherwise
st_green -> 1 - green party makes statement(s) on environmental protection; 0 - otherwise
st_rrw -> 1 - radical right-wing party makes statement(s) on immigration/asylum, law and order, national way of life; 0 - otherwise

VARIABLES' DESCRIPTION IN self_clean:
See CCDP codebook, available here:
https://www.mzes.uni-mannheim.de/d7/en/datasets/comparative-campaign-dynamics-dataset

VARIABLES' DESCRIPTION IN alternative_rhetoric_codings.csv:
(see Appendix B for details)
cons_alt1 <- measure of campaign performance treating all justice statements as neutral
cons_alt1_lag <- measure of lagged campaign performance treating all justice statements as neutral
cons_alt2 <- measure of campaign performance restricting to 23 CMP issues
cons_alt2_lag <- measure of lagged campaign performance restricting to 23 CMP issues
cons_alt3 <- measure of campaign performance increasing left-wing issues
cons_alt3_lag <- measure of lagged campaign performance increasing left-wing issues

VARIABLES' DESCRIPTION IN all_polls.csv:
Country - CCDP country
Year - campaign year
Date - publication date
Party - party for which poll estimate is available
Estimate - estimate of voting intention
prevvote - vote share in previous general election
elecvote - vote share in this year's election